The SAVE America act:
really stupid or just kinda stupid?
I know this has been written about a million times; Brennan center for Justice has weighed in (against, natch) and Heritage is for it, but their linked data shows that fewer than 2000 fraudulent votes have been cast in modern American history. Their fraud map shows that most states have significantly fewer than 100 proven fraudulent votes in modern history. They are completely silent as to why this proves there’s a problem with fraud in this country.
But it’s worth discussing because , by some polls 80% of the electorate support this intervention. Sorry to link such a disreputable source, but it summarizes the concept relatively well.
I get it. No one wants fraud. We all want a voting system where every person gets one and only one vote (At least that’s what I want, I’m not certain the Republicans are on board with that vision, but let’s assume the best of them). But It turns out that we’ve ALREADY gotten fraud under control. 31 in a billion feels like a reasonable fraud level to me. Literally no US election would be changed by eliminating that number of fraudulent votes. Now, I know what you’re thinking, (I want a sandwich, probably not) “Ryan, that was 2014 and maybe there wasn’t any fraud before 2016, but there’s a rapist who KNOWS that there’s tons of fraud and is ready to move heaven and earth to find and prosecute it. It’s Jeffry Epstein (no). But is was his more powerful friend, Donald who is really, really really trying to find that fraud.
Has he? With all the resources of the Justice (sic) Department?
Let’s find out.
That seems right, but let’s see if there’s a better source than young Tuvok.
How about Noted phallic rocket Billionaire’s Trump protection organization?
The Justice Department has struggled to meet White House demands to prosecute noncitizen voters as conspiracy theories that President Donald Trump and his allies have pushed in public fail to hold up legally.
So, no. Even with full control of the government for more than a year (after 4 years of, lets call it planning to do this? I don’t know his “mind”) they can’t find anyone to prosecute. I’ll let bipartisanpolicy.org bring this one home. They reported on three state level audits of the voter rolls. It’s pretty grim for the people counting on finding fraud.
Ohio Secretary of State Frank LaRose flagged 521 cases of potential noncitizen voting between 2019 and September 2023. But an Ohio Capital Journal review found just one case in which a noncitizen was ultimately charged with voter fraud. The review noted that this was due to a combination of accidental registration and a lag in how the department of motor vehicles updates an individual’s citizenship status after naturalization.
Utah performed a citizenship review of its entire voter registration list from April 2025 through January 2026. After a time-intensive, multi-step review of more than 2 million registered voters, they identified only one confirmed instance of noncitizen registration and zero instances of noncitizen voting.
Georgia: In October 2024, Georgia Secretary of State Brad Ratffensperger announced that a citizenship audit of 8.2 million registered voters identified 20 noncitizens who had registered to vote. Of those 20 individuals, just nine had cast ballots in prior elections, most before Georgia implemented enhanced citizenship verification through the Department of Driver Services in 2012. The cases were referred to local prosecutors for investigation.
Single digit fraudulent votes over multiple YEARS. Extrapolate that out and you’ve got maybe 50 cases of voter fraud in America per presidential cycle.
Now, let’s keep our heads about us. Heritage looks at that data and says this.
Oh!! HUNDREDS you say? Given that they have entries in their voter fraud map that go back to 1948 I’d HOPE there were hundreds of convictions. We se that in 2020 there were about 157 million votes cast for president.
Do you get the issue here? That’s a LOT of votes. Even in 1948 there were 54 million votes cast for president. They claim just over 1000. Since 1948.
This is just deeply dishonest. If there were just 2000 convictions for robbery since 1938 would we think robbery was a big problem? We would not.
I mean, the closest Presidential election of my lifetime was 2020. Bush “won” by 537 votes (we assume, not certain what would have happened if the recounts went ahead as planned.) so if we take all the proven fraudulent votes from the last 80 years and concentrate them in one state in one election they can change that one election. By the same minuscule margin it was won by. Doesn’t seem to me that that rises to a level that we need to care about.
That brings us to House races. I looked through the MIT electionlab database and found that there were 42 races decided by 400 or fewer votes (since 1929 when house districts were made the same size). 22 of those were won by Republicans. So if you take the 9 cases of non-citizen voting fraud in Georgia and just arbitrarily decide that only 2% of cases were found AND that all those cases were in every house district since 1948 then you could possibly have flipped as many as 20 house seats to Democrat hands. That 400 per house district number results would require more than 2 million fraudulent votes across the nation every year for the last 78 years. So rather than the hundreds we know about we’d need 156 million. Is it probable that we’re missing 2 million fraudulent votes cast every year? Seems REALLY unlikely to me. That would mean that for every fraudulent vote we find there are at least 100,000 in the walls. And, last I checked, fraudulent votes were not rats.
That seems like a lot to miss, but we know the Justice (sic) department and local police are sometimes pretty inept (or asked to fail intentionally) but let’s compare it to robberies. It looks like maybe a third of robberies are cleared (reported to the cops and then solved) and fewer than half are reported. So if you report half and solve a third then you see that convictions are about a 6th of the robberies. If we keep that rate then the 9 cases in Georgia become 54. To get to 400 you’d need to have missed 98% of the fraud. I’ve seen no evidence ANYWHERE that ANYONE thinks that’s the right ratio. And that 400 is in a state with 14 congressional districts. so, to flip those 20 historical elections you’d need to assume that you’re only catching about 0.16% of the fraud cases. That’s BONKERS, can’t possibly be true. I get that a robbery has a victim, but given the amount of heat that’s currently being heaped on the Justice (sic) department I can’t see that it’s possible that were only finding a tenth of a percent of the fraud out there. It’s not as if the investigations are terribly complex. Just compare the voting rolls with a more fully documented database. DOGE could have done it for every state if they’d cared to do something useful with their IT prowess rather than harmful nonsense.
I know that that was a lot of numbers. I know it was a lot of details. And you’re probably saying to me, “Ryan, why did you tie yourself in knots showing that there’s very little fraud? You know that the Republicans will just say “Man, I know you think it’s not a problem, but it’s no big deal to show proof of citizenship, so why not make the system more secure? Even 9 cases in a state are too many”
I did it because it’s important. The SAVE act, as written, is likely to disenfranchise millions of voters . Not everyone has a birth certificate handy (especially if you’ve got a non-enhanced license that you’ve used as ID for decades. And/or oyu’re mari) and tons of people don’t have a passport. None of those things are free. I get it, you think CAP is probably overreacting and it won’t disenfranchise millions. Cool, let’s assume they are 2 orders of magnitude off. It’ll just disenfranchise tens of thousands of people. Thats tens of thousands of people every election that were entitled to vote, but couldn’t becasuew of this new fraud prevention scheme.
Seems bad, but not TOO bad, that’s a number that’s unlikely to flip many elections, but it’s certainly a lot of people. 20,000 people in the country would average out to 400 per state. again, not enough to flip may elections, but if we look at the facts we see that there are only single digit fraud cases per year. So that means that even if we assume we’re only catching 1% of fraud cases now we still would be disenfranchising more eligible voters than there are fraudulent votes.
Super dumb.
And I won’t even get to the fact that those disenfranchised voters are likely to be mostly Republican . Maybe don’t mention it to them. They don’t really like facts anyway.